I was looking at national savings figures for the United States when the Australian National Accounts came out yesterday, which is why I belatedly noticed the negative households savings figure.

The US has also experienced a big decline in household savings, but they remain positive at around 3 per cent of GDP. Retained corporate earnings are between 0 and 2 per cent of GDP depending on how you measure depreciation. These small positive contributions are wiped out by the government budget deficit (around 5 per cent for the Federal government – the states are also in deficit, but I don’t have a number yet). More on all this is available from the Bureau of Economic Analysis.

One interpretation of all this is that people from outside the US (and, for that matter) Australia, are eager to buy US assets, and Americans are simply cashing in the consumption benefits. I don’t agree. A steady decline in household savings seems to be occurring wherever financial markets have been liberalised. At the moment, the whole system is being kept in balance by massive purchases of US dollars by Asian central banks, but this can’t continue indefinitely.

It’s therefore time to invoke Stein’s Law – if a process can’t continue indefinitely, it won’t. There seems no prospect of an exogenous shift in the behavior of households or of a return to fiscal probity by the US government. I conclude that a return to equilibrium must involve an increase in real and nominal interest rates, probably facilitated by inflation. Even allowing for an inflationary cushion, this will not be a pleasant process for heavily indebted Australian households. American householders are protected by the structure of mortgage contracts, which allows them to lock in low rates, but the costs will be borne elsewhere in the financial system.

Apart from showing near-zero GDP growth for the quarter, the latest national accounts released today by the Australian Bureau of Statistics include the startling (to me, anyway) information that Australia now has negative household savings. I’ve reproduced the relevant bit of the release below.

Household saving ratio

In both trend and seasonally adjusted terms the household saving ratio was negative in the June quarter 2003 implying that household consumption was greater than household disposable income. In trend terms the ratio was -1.2% in the June quarter and in seasonally adjusted terms it was -1.3%. The deterioration in the saving ratio in recent quarters has been driven by both a slow down in the rate of growth of disposable income and the continued strength of household consumption expenditure. The movement in disposable income has been affected by the very weak income results for the farm sector arising from the drought. The impact occurs because the household sector defined in the national accounts includes unincorporated businesses and therefore includes most farm businesses. Consequently, most farm income (included as a significant component of ‘gross mixed income’ ) is also part of total household income. Although seasonally adjusted household saving has been negative in the past three quarters, net national saving has been positive over the same period. The net national saving ratio in the June quarter was 2.5% in seasonally adjusted terms.

Caution should be exercised in interpreting the household saving ratio in recent years, because major components of household income and expenditure may still be subject to significant revisions. The impact of these revisions on the saving ratio can cause changes in the apparent direction of the trend. The following graph presents the household saving ratio derived from trend and seasonally adjusted data (see Explanatory Notes).

As the graph shows, although the latest figures may be distorted by the drought etc., the long-term trend has been clearly negative, and the decline goes back further than this (the Fitzgerald report on declining national savings was commissioned at the beginning of this period.

When I responded to the Fitzgerald report, I argued that it was misleading because it failed to take account of investment in human capital. But we’ve done miserably on this score in the last decade or so, with school completion rates declining in the early 90s (they’ve since recovered a bit) and domestic higher education commencements frozen since 1996. In both cases, there was a direct link to expenditure cuts imposed in the name of economic efficiency.

I haven’t yet managed to work through to an aggregate national savings figure. But with the Federal government budget roughly balanced in accrual terms, the contribution from government savings can’t be large, and I’d be surprised if retained earnings of corporations accruing to Australian owners amounted to more than 3 or 4 per cent of GDP. So this suggests that Australian national savings are approximately zero, or in other words, that all net investment in Australia must now be financed by foreign debt or equity investment.

One reason for this negative saving is the fact that, thanks to the property bubble, people can spend more than they earn and still, apparently, get richer. But there’s a fallacy of composition here. We can’t all sell our houses to cash in this wealth – if we did, prices would fall and the wealth would disappear.

Of course, if we could persuade some overseas buyers to purchase a million or so houses at current prices, our problems with foreign debt would be over. But although it’s not precisely true that the only potential buyers of Australian houses are Australian residents, it’s a good enough approximation for economic analysis. A few thousand wealthy HongKongers may want a Sydney bolthole, and there are probably a few thousand more footloose global professionals in the market, but not enough, I think, to make a real difference.

Update My wife Nancy, who’s paying more attention than I am, tells me there’s nothing new in the negative household savings story, which is confirmed by a look at the graph (savings have been negative for three or four quarters now) and a quick Google. As so often, I’m a bit behind the times, but I’m still surprised there hasn’t been more comment on this.